Longreads + Open Thread

Quick note to readers: we're now offering a better RSS experience, including full-text RSS feeds of paywalled posts for paying subscribers. You can access your unique RSS link here.

Longreads

• Quantian has a mathematical model for why it's good idea to cut losses rather than averaging down, at least in a case where your confidence relative to the market is unchanged. It's a necessarily stylized example, and there are cases where averaging down works well—specifically when there's a catalyst that some investors read as negative but that you read as positive or vice-versa. There are other, more psychological reasons ($, The Diff) , so it's nice to know that there's a mathematical justification, too.
• Felix Stocker on State Farm's success in insurance. Insurance is an interesting sector where some of the largest players are owned by policyholders rather than entirely by outside investors. In other industries, like retail and banking, co-ops exist, but they're rarer among the largest companies. Part of this is a story of exploiting preexisting infrastructure and selection effects: the company used data from an industry association to identify the market opportunity, and originally only underwrote policies for people who were "members of Farm Bureaus, Farm Mutual Insurance Companies, their immediate families and those eligible for membership in such organisations." The tradeoff for any growth company in an established industry is that there will be a visible path to growth using what already exists—but adopting all of it necessarily means becoming one of the stodgy incumbents the business was going after in the first place.
• Ian Hickson reflects on eighteen years at Google. A good contribution to the debate over how cynical established companies are when they try to set industry standards: "My mandate was to do the best thing for the web, as whatever was good for the web would be good for Google (I was explicitly told to ignore Google's interests)." (Emphasis added.) The piece also has good reflections on corporate decisions and incentives: after layoffs, he notes that people were less willing to take risks, because they're optimizing for staying above the layoff cutoff. Paradoxically, a strategy of more frequently culling more deeply might have the opposite effect on risk tolerance, if the threshold is set high enough that it's only by taking risk that you can do well enough to stay.
• Dwarkesh Patel interviews historian Andrew Roberts, author of, among other things, a wonderful biography of Napoleon (briefly noted in The Diff here). Much to mull in here, including Dwarkesh's observation that the Clausewitz rule-of-thumb that it takes a 3:1 ratio for attackers to overwhelm defenders held roughly true for a surprisingly long time and Roberts' claim that a modern Napoleon would move to Silicon Valley and start a tech company.
• Razib Khan in Palladium on the wonders of modern genomics. Sequencing the human genome has amazing applications in the future, but it's also affected how we see the past: we can get a more detailed look at the nature and timing of migration and conquest now that we know it leaves a genetic signature.
• In this week's Capital Gains, we look at why counterparty risk matters, in a micro and macro sense. On a micro level, the more life-changing the proceeds of a bet are, the more you have to consider the probability that you'll be repaid. On a macro level, modern monetary systems mostly exchange references to currency rather than currency itself, and that means that any default anywhere in the system ripples outward.
• This week's episode of our new podcast, The Riff, is a discussion of OpenAI, big tech's incumbency advantage, and which big tech company I'd short if I had to. Listen: YouTube/Apple/Spotify.

Books

• eBoys: The First Inside Account of Venture Capitalists at Work: A concept fiction authors sometimes have fun with is telling a story in which the protagonist gradually goes insane. This can be very evocative, albeit uncomfortable. In eBoys, Randall Stross writes about the team at Benchmark who, in the 90s, made what was then the best venture investment of all time ($7m into eBay in 1997, a stake that was worth $4.2bn in 1999). But they missed Priceline, prudently noting that the company managed to combine high marketing expense relative to sales, negative gross margins, and a supply problem. The book shows Benchmark's thinking evolve towards extreme optimism over just a few quarters (at one point, they're talking about how good a company is as a strategic acquisition—when the company in question is barely past the idea stage). Another benefit of the book is looking at how much valuations have changed: eBay's initial pricing was absurdly cheap, but at around the same time, VCs were putting a high valuation on consulting companies like Scient. At one point, they fund a security company whose claims struck me as obviously wrong. Which raised the question of why the VCs wouldn't know to ask more questions about the technical aspects of what they were buying—but later developments in the book indicate that the reason the world has more ambient knowledge about the limits of cryptography is that bloggers publicly attacked security snake oil when it was fundable in the 90s. It's a good reminder that, when people are reading about the 2020s some time in the 2050s, there will be a lot that should have been obvious to us today, but that we were totally oblivious to.

Open Thread

• Drop in any links or comments of interest to Diff readers.
• Is there any good account of the contemporary VC investment process, similar to eBoys? Sometimes, the world really lucks out and a business will give a writer a surprising level of access; have we been that lucky this cycle, too?
• We're interested in talking to buy-side investors involved in the chemicals space, with a focus on what kinds of alternative data you use and what you use it for. If you fit that description (or know someone who does), we'd love to chat! (Off the record by default here.)

Diff Jobs

Companies in the Diff network are actively looking for talent. A sampling of current open roles:

• A company building the new pension of the 21st century and building universal basic capital is looking for a product designer with fintech experience. (NYC)
• A diversified prop trading firm with a uniquely collaborative team structure is looking for experienced software engineers. (Singapore or Austin, TX preferred)
• A data consultancy is looking for a senior data scientist with prior experience in marketing data science and e-commerce. (NYC)
• A startup building a new financial market within a multi-trillion dollar asset class is looking for a junior ML engineer, especially someone interested in using LLMs to make unstructured data more tractable. (US, Remote)
• The leading provider of advanced options analytics — “the ASML of options trading” — is growing rapidly, very profitable, and looking for a generalist who can excel in chief of staff and business development functions. A trading, quant, or similarly technical background is a big plus. (Connecticut, NYC)

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.